surface export of gapdh/sdh, a glycolytic … glucose molecule. gas metabolizes glucose and derives...

Surface Export of GAPDH/SDH, a Glycolytic Enzyme, Is Essential forStreptococcus pyogenes Virulence

Hong Jin, Shivangi Agarwal, Shivani Agarwal, and Vijay Pancholi

Department of Pathology, Ohio State University College of Medicine, Columbus, Ohio, USA

ABSTRACT Streptococcal surface dehydrogenase (SDH) (glyceraldehyde-3-phosphate dehydrogenase [GAPDH]) is an anchorlessmajor multifunctional surface protein in group A Streptococcus (GAS) with the ability to bind important mammalian proteins,including plasmin(ogen). Although several biological properties of SDH are suggestive of its possible role in GAS virulence, itsdirect role in GAS pathogenesis has not been ascertained because it is essential for GAS survival. Thus, it has remained enigmaticas to “how and why” SDH/GAPDH is exported onto the bacterial surface. The present investigation highlights “why” SDH isexported onto the GAS surface. Differential microarray-based genome-wide transcript abundance analysis was carried out usinga specific mutant, which was created by inserting a hydrophobic tail at the C-terminal end of SDH (M1-SDHHBtail) and thus pre-venting its exportation onto the GAS surface. This analysis revealed downregulation of the majority of genes involved in GASvirulence and genes belonging to carbohydrate and amino acid metabolism and upregulation of those related to lipid metabo-lism. The complete attenuation of this mutant for virulence in the mouse model and the decreased and increased virulence of thewild-type and mutant strains postcomplementation with SDHHBtail and SDH, respectively, indicated that the SDH surface exportindeed regulates GAS virulence. M1-SDHHBtail also displayed unaltered growth patterns, increased intracellular ATP concentra-tion and Hpr double phosphorylation, and significantly reduced pH tolerance, streptolysin S, and SpeB activities. These pheno-typic and physiological changes observed in the mutant despite the unaltered expression levels of established transcriptionalregulators further highlight the fact that SDH interfaces with many regulators and its surface exportation is essential for GASvirulence.

IMPORTANCE Streptococcal surface dehydrogenase (SDH), a classical anchorless cytoplasmically localized glycolytic enzyme, isexported onto the group A Streptococcus (GAS) surface through a hitherto unknown mechanism(s). It has not been known whyGAS or other prokaryotes should export this protein onto the surface. By genetic manipulations, we created a novel GAS mutantstrain expressing SDH with a 12-amino-acid hydrophobic tail at its C-terminal end and thus were able to prevent its surface ex-portation without altering its enzymatic activity or growth pattern. Interestingly, the mutant was completely attenuated for viru-lence in a mouse peritonitis model. The global gene expression profiles of this mutant reveal that the surface exportation of SDHis mandatory to maintain GAS virulence. The ability of GAS as a successful pathogen to localize SDH in the cytoplasm as well ason the surface is physiologically relevant and dynamically obligatory to fine-tune the functions of many transcriptional regula-tors and also to exploit its virulence properties for infection.

Received 5 April 2011 Accepted 29 April 2011 Published 31 May 2011

Citation Jin H, Agarwal S, Agarwal S, Pancholi V. 2011. Surface export of GAPDH/SDH, a glycolytic enzyme, is essential for Streptococcus pyogenes virulence. mBio 2(3):e00068-11. doi:10.1128/mBio.00068-11.

Editor Rino Rappuoli, Novartis Vaccines and Diagnostics

Copyright © 2011 Hong et al. This is an open-access article distributed under the terms of the Creative Commons Attribution-Noncommercial-Share Alike 3.0 UnportedLicense, which permits unrestricted noncommercial use, distribution, and reproduction in any medium, provided the original author and source are credited.

Address correspondence to Vijay Pancholi, [email protected].

Streptococcus pyogenes (group A Streptococcus [GAS]) is the hu-man pathogen that causes the widest variety of diseases, rang-

ing from mild pharyngitis and impetigo to severe and often fataltoxic shock syndrome, and it also causes autoimmune heart andkidney diseases as poststreptococcal sequelae (1). Although GAS isknown to cause primarily localized, noninvasive, and mild infec-tions, the invasive and more-severe GAS infections are not un-common as there are 10,000 cases of invasive GAS disease in theUnited States and over 500,000 GAS infection-related deaths peryear worldwide (2). Despite the availability of sequence informa-tion for several GAS genomes and detailed characterization oftheir virulence factors, the pathogenic mechanisms of GAS stillremain elusive (1). Therefore, elucidation of precise mechanisms

underlying GAS pathogenesis is expected to facilitate develop-ment of effective therapeutics against S. pyogenes.

GAS has developed complex gene regulatory networks to senseand respond to subtle environmental changes (3–5) in order tosurvive in their intricate and continuously changing habitatswithin the human body and mucosal surfaces. Regulation of vir-ulence genes and GAS pathogenesis are primarily dependent onthese networks (4, 6–8). Many of the genes belonging to thesenetworks also interface with GAS metabolism or nutrient trans-port and regulation of catabolic-controlled protein (CcpA) genes(5, 9–11).

The glycolytic pathway is a universal and conserved metabolicpathway through which two energy-rich ATPs are produced from

RESEARCH ARTICLE

May/June 2011 Volume 2 Issue 3 e00068-11 ® mbio.asm.org 1

on July 24, 2019 by guesthttp://m

bio.asm.org/

Dow

nloaded from

mbio.asm.org

http://mbio.asm.org/

one glucose molecule. GAS metabolizes glucose and derives en-ergy from the glycolytic pathway for its growth and survival. Glyc-eraldehyde 3-phosphate dehydrogenase (GAPDH) is the key gly-colytic enzyme that catalyzes the oxidation and phosphorylationof glyceraldehyde-3-phosphate to an energy-rich intermediate,1,3-bisphosphoglycerate, in the presence of NAD as a cofactor.Typically, GAPDH is a cytoplasmic protein lacking a signal se-quence or a hydrophobic anchor. Since our first report on theidentification of GAPDH as a major surface protein of S. pyogenes(and hence termed the streptococcal surface GAPDH or strepto-coccal surface dehydrogenase [SDH]/Plr/SPy0274) (12), severalreports have demonstrated that GAPDH is either expressed on thecell surface or secreted in numerous Gram-positive and Gram-negative bacteria, fungi, and parasites, including a bioterror agent,Bacillus anthracis (13–15). SDH has been categorized as an an-chorless bacterial surface protein (12, 14). Various nonglycolyticfunctions of SDH, such as auto-ADP-ribosylation (16), ability tobind to mammalian structural proteins (fibronectin, laminin, my-osin, actin, lysozyme) (12) and to proteins belonging to the hu-man fibrinolytic system (plasmin [12, 17, 18] and uPAR [19]), andability to regulate host cell signaling (20) indicate that SDH playsan important role in GAS virulence. Similarly, the ability of Liste-ria monocytogenes GAPDH to ADP-ribosylate Rab5a and subse-quently impair host phagolysosome (21) and the recently re-ported antiphagocytic activity of surface GAPDH/SDH (22)indicate that GAS SDH/GAPDH can indeed serve as a virulencefactor. Moreover, the ability of anti-SDH/GAPDH antibody tomediate opsonophagocytosis of GAS (19) and to provide protec-tion against GAS challenge (V. Pancholi, unpublished report), inconjunction with similar reports on other pathogens such as Bru-cella abortus (23), Edwardsiella tarda, (24), Streptococcus agalactiae(25), and B. anthracis (15), indicate that GAPDH in general canserve as a potential vaccine target.

However, it is still unclear (i) how and why SDH/GAPDH isexported onto the bacterial surface, and (ii) how SDH participatesin GAS pathogenesis. SDH is encoded by a single gene and hence,being indispensable, our several attempts to create a GAS mutantlacking SDH were unsuccessful. Previously, by exploiting a novelstrategy, we created a GAS mutant (M1-SDHHBtail) expressingSDH with a hydrophobic tail at its C terminus (18) that did notexport SDH (SDHHBtail) onto the cell surface and was retainedexclusively within the cytoplasm. The growth pattern andGAPDH activity of this mutant were comparable to those of thewild-type strain. Interestingly, the mutant demonstrated im-paired antiphagocytic activity and reduced ability to adhere tohuman pharyngeal cells (18), which are considered the two majordeterminants of virulence in GAS. These findings thus stronglyindicate that there is functional relevance for the exportation ofanchorless surface SDH onto the GAS cell surface. The presentstudy, therefore, was carried out to address why SDH/GAPDH isexported onto GAS and other bacterial surfaces and to elucidate itsrole in GAS pathogenesis.

RESULTSPrevention of SDH export in the M1-SDHHBtail mutant leads tosubstantial changes in gene regulation. To determine the effectof inhibition of surface exportation of SDH, microarray analysiswas performed employing the wild-type M1-SF370 (M1-WT) andisogenic mutant (M1-SDHHBtail) GAS strains grown to late logphase. The global gene expression profile of M1-SDHHBtail in

comparison to M1-WT revealed significant (�2 fold; P � 0.05)changes in the transcript abundance of 192 (10.6%) of the totalreported genes in the M1-SF370 genome (26). Among the differ-entially regulated genes in M1-SDHHBtail, 128 genes (66.7%) weredownregulated, while the other 64 genes (33.3%) were upregu-lated (Table 1). Notably, about 46% of the differentially expressedgenes (89 of 192 genes) were related to bacterial metabolism (Ta-ble 1), most of which belonged to carbohydrate metabolism (37genes), followed by energy production and conversion (14 genes),lipid metabolism (12 genes), and amino acid transport and me-tabolism (11 genes). Except for the genes involved in lipid andnucleotide metabolism, a substantial decrease in the transcriptabundance was observed for the genes belonging to general me-tabolism and nutrient transport. Additionally, 18 (9.37%) of thetotal 192 genes displaying differential regulation were found to beassociated with the virulence. Transcript abundance of ~72.2%(13/18 genes) of these genes was found to be significantly de-creased, indicating that the prevention of the SDH exportation oraccumulation of SDH within the cytoplasm (18) adversely affectsstreptococcal metabolism and virulence. Apart from validatingthe data obtained from microarray analysis with real-time quan-titative reverse transcription-PCR (qRT-PCR) (see Table S3 andTable S4 in the supplemental material), we also studied severalcorresponding biochemical functions to establish a functionalcorrelation with the observed transcript abundance.

Prevention of SDH export adversely affects carbohydratetransport and metabolism. Prevention of SDH exportation re-sulted in downregulation of 36 out of 37 differentially regulatedgenes in the carbohydrate transport and metabolism category inthe M1-SDHHBtail mutant strain as revealed by the comparativemicroarray and qRT-PCR-based analyses (see Table S3 and Ta-ble S4 in the supplemental material). Notably, this includeddownregulation of 11 out of 18 predicted or identified genes spe-cific to carbohydrate phosphotransferase (PTS) systems/ABCtransporters (Table S5). In particular, multiple operons (SPy1057-1060, SPy1294-1296, SPy1299-1301, SPy1306, and SPy1986) con-taining genes for utilization/transport of nonglucose substrates,such as maltose/maltodextrin, mannose, mannose/fructose,N-acetylneuraminate, �-glucoside, fructose, cellobiose, galactose,lactose, and trehalose were downregulated (8- to 32-fold or 2 to 5log2 units). We also observed a 2-fold decrease in the SPy1986transcript (with 90% confidence limit), which was originally an-notated as glucose-specific PTS system and recently has been an-notated as MalT (maltose transporter) in a type M1 GAS strain,MGAS_5005 (27).

Since about half of the differentially regulated genes in theM1-SDHHBtail mutant were related to metabolism, most of whichwere downregulated and essentially belonged to carbohydratetransporters, we investigated the growth properties of the M1-SDHHBtail mutant and the wild-type GAS strains in chemicallydefined medium (CDM) supplemented with different sugars asthe sole carbon source (Fig. 1). Our results revealed that the lack ofsurface exportation of SDH did not change the growth character-istics of the mutant in complex nutritionally enriched THY me-dium (Todd-Hewitt broth supplemented with 0.5% yeast ex-tract). However, when the mutant was grown in CDM withdifferent sugars as the sole carbon source, different growth pat-terns of M1-SDHHBtail were observed compared to M1-WT. Moreimportantly, the growth rates of the M1-SDHHBtail mutant werenot affected when grown in CDM with glucose, sucrose, fructose,

Jin et al.

2 ® mbio.asm.org May/June 2011 Volume 2 Issue 3 e00068-11


bio.asm.org/

Dow

nloaded from

mbio.asm.org


and mannose (Fig. 1). On the other hand, when maltose or malto-dextrin was used as the sole carbon source, the growth rate of theM1-SDHHBtail mutant was reduced (Fig. 1), indicating that theinhibition of SDH exportation or increased intracellular concen-tration of SDH directly influences the ability of GAS to metabolizeand transport specific carbon sources.

Cytoplasmic retention of SDH results in increased intracel-lular ATP concentrations and decreased acid tolerance. Mi-croarray analysis of the M1-SDHHBtail mutant revealed downregu-lation of 11 genes (SPy0148 to SPy0151, SPy0154, SPy0155,SPy0157, SPy0414, SPy0739, SPy1128, and SPy1849) and upregu-lation of three genes (SPy0755, SPy0757, and SPy0759) related toenergy production and conversion (see Table S3 in the supple-mental material). Since most of the downregulated genes be-longed to V-type Na� ATPase and upregulated genes belonged toproton-translocating ATPase, we hypothesized that the down-regulation of ATPase may result in a surplus intracellular concen-tration of ATP. To validate the microarray data, we measured theintracellular concentration of ATP in the M1-SDHHBtail mutantand wild-type strains by luciferin-luciferase assay. An 8-fold in-crease in the intracellular ATP concentration in the mutant(Fig. 2A) indicated that the prevention of SDH export and itssubsequent increase in the cytosol positively regulate the levels ofATP in the M1-SDHHBtail mutant.

The results described above also indicated that increased intra-cellular concentration of SDH may impair the ability of GAS toexpel H� due to the significantly reduced expression ofV1V0ATPases, resulting in reduced acid tolerance. Argininedeaminase system (ADS) has been previously reported to conferacid tolerance in GAS (28) and influence its ability to adhere andinvade host epithelial cells (29). We observed a significant down-regulation of SPy1541-SPy1549 genes (~32-fold) encoding en-zymes in the arginine deaminase system (ADS) of S. pyogenes (Ta-ble 1; also see Table S3 in the supplemental material). To validatethese results, we analyzed the ability of the wild-type and mutantGAS strains to survive and grow in the acidic environment bymonitoring their growth patterns in THY broth at pHs rangingfrom 4.0 to 7.5 (Fig. 2B). The M1-SDHHBtail mutant showed sig-nificantly retarded growth (P � 0.001) at pHs ranging from 4.5 to6.0 in comparison to the wild-type strain (Fig. 2B), indicating thatinhibition of SDH export onto the cell surface significantly af-fected the intracellular proton homeostasis in GAS. Thus, preven-tion of SDH exportation and the concomitant increase in its in-tracellular concentration adversely affect the ability of the GASmutant to survive in the acidic environment.

Inhibition of SDH exportation affects phosphorylation ofHPr, a central regulatory protein of the carbohydrate phospho-transferase system. As described above, the prevention of SDH

TABLE 1 Microarray-based global gene expression profiles for the M1-SDHHBtail mutant strain versus the wild-type M1-SF370 strain showing thenumber of significantly differentiated genes belonging to different functional categories

Functional categoryof genes

Total no.of significantlydifferentiated genesa

%of significantlydifferentiated genes

No.of upregulatedgenes

No.of downregulatedgenes

Amino acid transportand metabolism

11 5.7 3 8

Carbohydrate transportand metabolism

37 19.3 1 36

Cell divisionand chromosome partition

1 0.5 1 0

Cell envelopeand biogenesis

2 1.04 0 2

Cell motilityand secretion

1 0.5 0 1

Coenzyme metabolism 3 1.6 0 3DNA replication

and recombination4 2.1 3 1

Energy productionand conversion

14 7.3 3 11

Function unknown 36 18.8 10 26General function

prediction only15 7.8 5 10

Inorganic iontransport

1 0.5 0 1

Lipid metabolism 12 6.3 11 1Nucleotide transport

and metabolism4 2.1 4 0

Phage 15 7.8 10 5Posttranslational modification 5 2.6 3 2Secondary

metabolite biosynthesis5 2.6 2 3

Transcription 5 2.6 1 4Translation

and ribosomal structure4 2.1 2 2

Virulence genes 18 9.4 5 13Total no. (%) 192 100 64 (33.3) 128 (66.7)a The total number of significantly differentiated genes for the different functional categories of genes and the totals are shown in boldface type.

SDH, a Virulence Factor-Regulating Virulence Factor



bio.asm.org/

Dow

nloaded from

mbio.asm.org


exportation resulted in downregulation of several genes related tocarbohydrate transport and metabolism. Many of these genes be-long to the phosphotransferase (PTS) system (see Table S5 in thesupplemental material). Bacterial systems have developed com-plex mechanisms to adapt to environmental conditions, such ascarbon catabolite repression (CCR), which facilitates utilizationof preferential substrates like glucose instead of other nonpre-ferred substrates when more than one carbohydrate exists in theenvironment (11). The phosphoenolpyruvate (PEP):carbohy-drate phosphotransferase system is a well-described regulatorysystem for CCR through which bacteria transport and phosphor-ylate the carbohydrates (30). In this study, our microarray analysisrevealed downregulation of several PTS-related genes without sig-nificantly altering the transcript abundance of the ccpA gene(SPy0514). This indicates that the observed differential expressionof PTS-related genes is likely due to an altered interaction betweenHpr and CcpA, which in turn is influenced by the phosphoryla-tion status of Hpr. Since the ATP levels were significantly elevatedin the SDH mutant (Fig. 2A), we speculated that the increasedintracellular concentration of SDH somehow modulates the Hprphosphorylation and/or function of CcpA.

To examine the relative changes in the phosphorylation statusof Hpr in the M1-SDHHBtail mutant strain in comparison to thatin the wild-type GAS strain, we determined expression levels ofdifferentially resolved phosphorylated forms of HPr by nativePAGE analysis. This analysis allowed us to differentiate between

fast-migrating Ser-phosphorylated HPr (HPr-Ser-P or HPr-P1)and/or His/Ser-phosphorylated HPr (HPr-His/Ser-P or HPr-P2).Parallel Western blot analysis of the whole-cell lysates of M1-SDHHBtail mutant and M1-WT strains was carried out to deter-mine relative endogenous protein levels of HPr and CcpA. Al-though the protein expression profiles of CcpA and Hpr werecomparable in both GAS strains (Fig. 3A), a significant increase inthe amount of doubly phosphorylated Hpr (Hpr-Ser-P, -His-P[HPr-P2]) in the M1-SDHHBtail mutant was observed (Fig. 3B).This indicated that the retention of SDH in the cytoplasm resultsin increased phosphorylation of Hpr possibly due to increasedATP production (Fig. 2A) as also reported previously (31).

Retention of SDH within the cytosol increases lipid/fattyacid biosynthesis. Unlike the expression profiles of carbohydratemetabolism genes in the M1-SDHHBtail mutant as describedabove, microarray-based differential gene expression analysis forthe mutant in comparison to the wild-type M1-WT GAS strainsrevealed 2- to 4-fold upregulation of 13 lipid biosynthesis geneslocated in a cluster (fabM/phaB/SPy1758, fabT/SPy1755, fabH/SPy1754, acpP/SPy1753, fabK/SPy1751, fadD/SPy1750, fabG/SPy1749, fabF/SPy1748, accB/SPy1747, fabZ/SPy1746, accC/SPy1745, accD/SPy1744, and accA/SPy1743) (see Table S3 in thesupplemental material). These genes collectively encode enzymesthat belong to the Gram-positive fatty acid synthesis type II(FASII) pathway involving a series of enzymatic steps (32). FASIIof Gram-positive bacteria is well characterized in the human

FIG 1 Growth curves of the wild-type M1-SF370 (M1-WT) and its isogenic mutant (M1-SDHHBtail) GAS strains in chemically defined medium supplementedwith glucose, sucrose, fructose, maltose, maltodextrin, and mannose. Error bars represent standard error of the mean OD600nm obtained from three independentexperiments.

Jin et al.



bio.asm.org/

Dow

nloaded from

mbio.asm.org


pathogen Streptococcus pneumoniae and is considered the univer-sal fatty acid synthesis pathway, since similar gene clusters are alsofound in Enterococcus, Clostridium, and Lactococcus species (32).Comparison of the fatty acid synthesis cluster of S. pyogenes withthat of S. pneumoniae revealed significant homology and similargenetic organization. We therefore predicted that the increasedtranscript abundance of fatty acid biosynthesis genes may result inthe increased fatty acid content in the M1-SDHHBtail mutant.

To validate these data, we analyzed thetotal fatty acid content and compositionalvariation in the M1-SDHHBtail mutantand wild-type strains by gas chromatog-raphy (Table 2). The significant increasein the total fatty acid content in the M1-SDHHBtail mutant (1.64 fold; P � 0.013)was in accordance with the increasedtranscript abundance of lipid/fatty acidbiosynthesis genes, SPy1743-SPy1758.The unsaturated fatty acid (UFA) and sat-urated fatty acid (SFA) profile in the mu-tant revealed significantly increased (P �0.03) synthesis of individual saturatedfatty acids (12:0, 14:0, 16:0, and 18:0) ex-cept those possessing carbon chainlengths of 10:0 (not shown for M1-WTand not detected in the M1-SDHHBtail

mutant) and 13:0. Also, the reduction inthe UFA/SFA (131.9:96.5) ratio from 1.37(M1-WT) to 1.02 (166.43:163.1 in M1-SDHHBtail) indicated an upregulation ofsaturated fatty acid content in the mutant(Table 2). Collectively, these results re-vealed that the prevention of SDH expor-tation directly or indirectly affects lipidbiosynthesis and fatty acid compositionin the GAS.

Prevention of SDH exportation downregulates major GASvirulence factors. Microarray analysis of the M1-SDHHBtail mu-tant revealed differential expression profiles for 33 of the mostimportant and well-characterized virulence genes (Table 3). Thelatter include adenine dinucleotide glycohydrolase (Nga/SPy0165),streptolysin O (Slo/SPy0167), C3 family ADP-ribosyltransferase(SpyA/SPy0428), exotoxins (SpeJ/SPy0436, SpeB/SPy2039),

streptolysin S (SagA operon/SPy0738-SPy00746), pullulanase (PulA/SPy1972),streptokinase A (Ska/SPy1979), collagen-like surface protein (Scl/SPy1983), inhib-itor of complement-mediated lysis (Sic/SPy2016), M protein (Emm1/SPy2018),mitogenic factor (mf/SPy2043) and cap-sule (Has operon/SPy2200-SPy2202).These gene products have been shown toplay an important role in S. pyogenes col-onization, persistence, dissemination,and proliferation (1, 27). Many of thesevirulence factors which are displayed assurface proteins contribute to the outer,electron-dense, fuzzy layer on the surfaceof GAS. The mutant (0.94 �m) was signif-icantly larger (34%) than the wild-typestrain (~0.72 �m). In the mutant, thisfuzzy layer was found to be absent, whichsuggests the loss of expression of severalsurface proteins, including M protein,which concurs with the microarray andqRT-PCR data (Fig. 4A; see also Table S4in the supplemental material).

We further validated the above results

FIG 2 (A) Intracellular ATP concentrations in the wild-type M1-SF370 (M1-WT) and its isogenicmutant (M1-SDHHBtail) GAS strains. ATP concentrations in samples were determined using luciferin-luciferase bioluminescence assay based on the standard curve obtained with known concentrations ofATP. Error bars represent standard errors of the mean ATP concentrations obtained from three inde-pendent experiments (triplicate wells for each experiment). (B) Effect of pH on growth of the wild-typeM1-SF370 (M1-WT) and mutant (M1-SDHHBtail) GAS strains. GAS strains were grown overnight inchemically defined medium (CDM), and culture density (600 nm) was determined as described inMaterials and Methods. Error bars represent standard error of the mean OD600nm obtained from threeindependent experiments.

FIG 3 Western blot analysis of the whole-cell lysate (CL) of the overnight-grown wild-type M1-SF370(M1-WT) and mutant (M1-SDHHBtail) GAS strains for the presence of SDH, CcpA, HPr, and CovR,using protein-specific antibodies. Equal amounts (50 �g total protein [25 ul] as the starting concentra-tion) of cell lysates of M1-WT and M1-SDHHBtail were used for the assay. (A) Western blot analysis ofequal volume (25ul) of the four serially 2-fold diluted samples of whole-cell lysates to determine theconcentrations of different proteins as indicated in the wild-type and mutant GAS strains. (B) Reactivityof mono- and doubly phosphorylated forms of HPr (HPr-P1 and HPr-P2) in the whole-cell lysates ofM1-WT and M1-SDHHBtail GAS strains.




bio.asm.org/

Dow

nloaded from

mbio.asm.org


by demonstrating the absence of the antiphagocytic M protein inthe cell wall extracts and the secretory streptococcal plasminogenactivator, Ska, in the culture supernatants of the M1-SDHHBtail

mutant, using anti-M-protein specific and anti-SKA antibodies,respectively, in Western blot analysis (Fig. 4B). Additionally, asubstantial decrease in the hyaluronic acid content in the M1-SDHHBtail mutant (20.82 � 1.79 �g/ml) compared to the wild-type strain (66.46 � 2.41 �g/ml) is in agreement with the observeddownregulation of genes belonging to the has operon in the M1-SDHHBtail mutant (Fig. 4C). These results corroborated our earlierreport showing significantly reduced survival of the M1-SDHHBtail

mutant in whole blood (18).Additionally, we also observed a 10- to 16-fold downregulation

of the streptococcal exotoxin-encoding genes of the sag operon(sagA-sagI, SPy0738-SPy0746) (33), which correlated with the 25-

fold decrease in streptolysin S (Sls) production in the mutant (100hemolytic units in M1-WT versus 3.98 hemolytic units in M1-SDHHBtail) (Fig. 4D). To rule out the possibility that the observedhemolysis was not due to streptolysin O (Slo), we analyzed thecontribution of Slo to hemolysis by the addition of trypan blue (anSls inhibitor) to the reaction mixtures. The resulting hemolyticratio, 0.0398 (M1-SDHHBtail/M1-WT) (not shown) indicated thatthe reduction in the hemolytic activity in the mutant was notcaused by Slo secretion but is essentially mediated by Sls (Fig. 4D).

Prevention of SDH export results in downregulation of SpeBand possibly its secretion. Streptococcal pyogenic exotoxin B(SpeB/SPy0274) is another virulence factor which degrades hostserum proteins, such as human extracellular matrix, immuno-globulins, complement components, and even S. pyogenes surfaceand secreted proteins (34). Downregulation of speB (7.01-folddownregulation found by microarray and 3.22-fold downregula-tion by qRT-PCR [P � 0.0001]; see Table S4 in the supplementalmaterial) in the M1-SDHHBtail mutant prompted us to comparethe SpeB-specific cysteine protease activity in the culture superna-tants of M1-SDHHBtail and M1-WT strains. This was accom-plished by measuring the fluorescence released upon casein hy-drolysis in the reaction mixture containing casein-fluoresceinisothiocyanate (FITC) (Fig. 5A). About 94% of the cysteine pro-tease activity observed in the culture supernatants of M1-WT wasinhibited by cysteine protease inhibitor E-64, indicating that mostof the detected hydrolytic activity in the supernatant was contrib-uted by cysteine protease (Fig. 5A). Compared to the wild type, theM1-SDHHBtail mutant displayed significantly reduced cysteineprotease activity (22% for M1-SDHHBtail and 100% for M1-WT)(Fig. 5A).

Similarly, SpeB-specific protease activity was also determinedby visualizing the zones of clearance due to casein digestion onmilk agar plates. The reduced size of the clearance zones aroundthe disk impregnated with the supernatant obtained from M1-SDHHBtail mutant strain on milk agar plates clearly demonstratedthat the retention of SDH in the cytoplasm significantly reducedSpeB activity in the mutant (Fig. 5B). Further, to determinewhether SpeB secretion is directly related to SDH surface expor-

TABLE 2 Fatty acid composition of the wild-type (M1-WT) andmutant (M1-SDHHBtail) GAS strainsa

Fatty acid peak

Fatty acid compositionb

P valueM1-WT M1-SDHHBtail

12:0 1.94 � 0.41 5.04 � 0.46 0.001613:0 1.10 � 0.21 1.37 � 0.17 0.092614:0 3.66 � 0.79 8.41 � 0.24 0.005016:1 w9cc 13.53 � 1.77 19.18 � 4.53 0.091716:1 w7c 21.21 � 4.14 31.27 � 7.04 0.061316:1 w5c 3.18 � 0.39 4.29 � 0.92 0.097816:0 71.58 � 9.37 109.79 � 9.54 0.007818:1 w9c 20.9 � 1.07 29.19 � 3.02 0.023318:1 w7c 66.88 � 1.16 74.76 � 9.52 0.145518:1 w5c 4.35 � 0.12 4.82 � 0.51 0.130018:0 18.23 � 1.54 38.49 � 7.37 0.021519:1 iso I 1.82 � 0.21 2.92 � 0.24 0.0049Total fatty acid 228.71 � 19.5 330.69 � 26.6 0.0064% fatty acid in cell mass 0.47 � 0.06 0.77 � 0.10 0.0129a The rows with significant differences as evaluated by one-tailed nonparametric t testwith Welch’s correction are shaded.b Mean � SD of three independent experiments. Shaded rows highlight significantincrease in specific fatty acid contents.c 16:1 w9c, fatty acid with 16-carbon chain with one double bond located at the ninthcarbon from the � (omega) end of the chain.

TABLE 3 Differentially expressed genes encoding virulence factors in the M1-SDHHBtail mutant straina

M1-SF370 ORF Gene Annotation Microarray median log2 ratio P value

SPy0165 nga Nicotine adenine dinucleotide glycohydrolase precursor �2.35 0.001SPy0167 slo Streptolysin O precursor �2.25 0.001SPy0428 spyA C3 family ADP-ribosyltransferase �1.29 0.003SPy0436 speJ Putative exotoxin (superantigen) �1.24 0.004SPy0738 to -746 sagA-sagI Streptolysin S-associated ORF �2.25 to �4.18 0.0002 to 3.72E�-06SPy1972 pulA Putative pullulanase �1.16 0.0238SPy1979 ska Streptokinase A precursor �1.50 0.0041SPy1983 scl Collagen-like surface protein �5.59 4.94E�09SPy1985 rgfB Exodeoxyribonuclease III �1.96 0.0044SPy2016 sic Inhibitor of complement-mediated lysis �2.13 4.76E�07SPy2018 emm1 M-protein type 1 �1.13 0.0121SPy2039 speB Pyrogenic exotoxin B �2.81 0.0007SPy2043 mf Mitogenic factor �1.83 0.0016SPy2200 to -2202 hasA-hasC Hyaluronate synthase, UDP-G-6-D, UDP-GPPIaseb �4.32 to�4.49 1.29 to 1.72 E�08SPy0711 speC Pyrogenic exotoxin C precursor, phage associated 1.54 0.0012Spy0712 mf2 Putative DNase, phage associated 1.69 0.0012Spy0737 epf Putative extracellular matrix-binding protein 3.45 0.0005Spy1357 grab Protein GRAB (protein G-related � 2M-binding protein) 2.42 5.45E�05a Shaded rows highlight significant up-regulation of specific genes in the mutant strain.b UDP-G-6-D, UDP-glucose-6-dehydrogenase; UDP-GPPlase, UDP-glucose pyrophosphorylase.

Jin et al.



bio.asm.org/

Dow

nloaded from

mbio.asm.org


tation, we similarly measured SpeB-associated cysteine proteaseactivity for the M1-WT strain complemented with the gene en-coding SDHHBtail. The significantly reduced size of the clearancezone around the disk impregnated with supernatant obtainedfrom the complemented strain in comparison to the wild-typestrain confirmed that SpeB secretion is directly related to SDHexport. Additionally, when the supernatants obtained from theM1-SDHHBtail mutant complemented with the gene encodingwild-type SDH was examined for SpeB activity, a further reduc-tion in the size of the clearance zone around the test disks wasobserved (Fig. 5B). These results also concurred with the anti-SpeB reactivity in the same volume of trichloroacetic acid (TCA)-precipitated supernatants obtained from the wild-type (M1-WT),mutant (M1-SDHHBtail), and their corresponding complementedstrains (Fig. 5C). These results indicate that the exportation ofSpeB is directly related to the surface export of SDH and that SDHmight bind to SpeB, facilitating its secretion. To determine theabove contention, we performed a blot overlay assay using culturesupernatants obtained from the M1-WT, M1-SDHHBtail, M1-WT::SDHHBtail, and M1-SDHHBtail::SDH strains. The direct bind-ing of SDH to the pro-SpeB molecule (~35-kDa protein) as re-vealed by anti-SDH reactivity after the blot was overlaid withpurified recombinant SDH indicated that SDH may potentiallyserve as a chaperone/carrier protein for SpeB, although SpeB itselfis a secretory protein (Fig. 5C). These results thus emphasize thesignificance of the process of surface export of SDH in terms offacilitating transport of other proteins or virulence factors which

may or may not possess classical surface export or secretion ma-chinery.

Surface export of SDH is essential for the maintenance ofGAS virulence. Microarray analysis in conjunction with otherfunctional analyses collectively indicated that the prevention ofSDH export onto the cell surface results in downregulation ofseveral virulence factors (Table 3). We therefore hypothesized thatthe inhibition of SDH export might adversely affects GAS viru-lence in an experimental mouse intraperitoneal infection model.A group of 20 mice injected with the M1-SDHHBtail strain dis-played 100% survival, while all 20 mice injected with the wild-typestrain died by day 3 postinfection (P � 0.0001) (Fig. 6A). Whilelive bacteria were recovered from the spleen (106 CFU/mg of tis-sue), lung, liver, and kidneys (5 � 104 to 8 � 104 CFU/mg oftissue) of mice infected with M1-WT, no bacteria were recoveredfrom the mice infected with the M1-SDHHBtail mutant, indicatingits attenuation for virulence. Together, these results indicate thatthe mere retention of SDH in the cytoplasm by preventing itsexportation onto the GAS surface adversely affect both GAS me-tabolism and virulence. Hence, the accumulation of SDH/GAPDH in the cytoplasm beyond its physiological concentrationresults in the attenuation of GAS virulence.

The above results raise an important question whether surfaceexportation of SDH per se is directly related to GAS virulence. Toaddress this question, we included an additional two groups ofmice (10 mice/group), one group infected with the M1-SDHHBtail

mutant strain complemented with the gene encoding wild-type

FIG 4 Effects of the cytoplasmic retention of SDH on the expression of virulence factors in GAS. (A) Transmission electron microscopy of the wild-type(M1-WT) and mutant (M1-SDHHBtail) GAS strains. (B) Relative expression levels of the M1 protein and streptokinase in the cell wall fractions and culturesupernatants of the M1-WT and M1-SDHHBtail GAS strains as determined by type M1-reactive 10B6 monoclonal antibody (Anti-M 10B6 Ab) and antistrepto-kinase antibody (Anti-SKA Ab) in Western blot analysis. (C) Hyaluronic acid contents in M1-WT and M1-SDHHBtail GAS strains. (D) Relative percent hemolysinactivity on sheep RBCs of the serially diluted (10-fold-diluted) culture supernatants of M1-WT and M1-SDHHBtail GAS strains. Lysis of RBCs in the undilutedculture supernatant of the M1-WT strain was set at 100%.




bio.asm.org/

Dow

nloaded from

mbio.asm.org


SDH and one group infected with theM1-WT strain complemented with thegene encoding SDHHBtail. In the firstgroup (M1-SDHHBtail::sdh), 30% micedied, indicating that the attenuated M1-SDHHBtail regained significant virulence(P � 0.0003) upon expression of wild-type SDH from the plasmid in conjunc-tion with the SDHHBtail from the genome(Fig. 6A). In the second group (M1-WT::sdhHBtail), 50% of the mice survivedthroughout a 10-day period in compari-son to no mice survived (100% death ob-served within 3 days) in the group infectedwith the wild-type (M1-WT) strain. Thisindicated that the virulence of the wild-typestrain was significantly reduced (P � 0143)as a result of the expression of SDHHBtail,which was not exported and was retainedsolely in the cytoplasm (Fig. 6A). Together,these results confirmed that the surface ex-port of SDH is essential to maintain GASvirulence and overaccumulation of SDH bypreventing its transport results in attenua-tion of virulence.

To further delineate the phenomenonof the surface export-related virulenceregulation in GAS, we compared the ex-pression levels of certain important genes,including those related to virulence inboth of the aforementioned comple-mented strains. Although several geneswere downregulated in the M1-SDHHBtail

mutant upon complementation with thegene encoding SDH, 5 out of 10 genesshowed relative upregulation, althoughthey did not achieve the wild-type expres-sion levels (Fig. 6B; also see Table S6 in thesupplemental material) possibly becauseof the expression of cytoplasmically re-tained SDHHBtail in the background. Thelatter trend may be responsible for the in-crease in virulence (Fig. 6A). Similarly,downregulation of all 10 genes in M1-WT::sdhHBtail mimicking the gene expres-sion profile of the M1-SDHHBtail con-curred with its significantly decreasedvirulence, although it did not becomeavirulent because of the presence of SDHcapable of being exported onto the sur-face.

Collectively, our data highlight the sig-nificance of the SDH export process per sein controlling various aspects of metabo-lism and virulence maintenance in GAS.

DISCUSSION

GAPDH is one of the key cytoplasmic en-zymes involved in glycolysis. As much asthe surface exportation of GAPDH in the

FIG 5 Relationship of the surface export of SDH and SpeB secretion. (A) Determination of the relativecysteine protease activities of SpeB present in the culture supernatants of M1-WT and M1-SDHHBtail

GAS using FITC-labeled casein. The released fluorescence activity by the culture supernatant of thewild-type GAS strain was set at 100% activity, and its specificity was determined in the presence ofcysteine protease inhibitor E-64. Values are means plus standard errors of the means (error bars)of three to six independent experiments. (B) Determination of SpeB-specific cysteine protease ac-tivity (hydrolysis of milk casein) present in the culture supernatants of M1-WT, M1-SDHHBtail,M1-SDHHBtail::sdh, and M1-WT::sdhHBtail GAS strains using milk agar. (C) Western blot analysis show-ing the effect on the secretion of SpeB in the culture supernatants of M1-WT, M1-SDHHBtail,M1-SDHHBtail::sdh, and M1-WT::sdhHBtail strains and the ability of SDH to bind secreted SpeB asdetermined by the blot overlay method.

FIG 6 Retention of SDH in the cytoplasm attenuates GAS virulence in an experimental mouseintraperitoneal infection model. (A) Survival/mortality curves for M1-WT and M1-SDHHBtail (20 miceper group) and their corresponding complemented strains (10 mice per group), M1-SDHHBtail::sdh andM1-WT::sdhHBTail (complemented strains created with pDC123 plasmid containing genes encodingSDHHBtail and SDH, respectively). Mice infected with GAS strains were monitored for 10 days postin-fection (P.I.) and statistically evaluated by the log rank test, and the results were plotted using GraphPadPrism 4 software. All the mock-infected mice survived throughout the observation period (not shown).(B) qRT-PCR-based expression analysis of the indicated genes in the complemented strains. The rela-tive fold change in the expression level was calculated with respect to the expression levels in thecorresponding parent strains (M1-SDHHBtail::sdh versus M1-SDHHBtail and M1-WT::sdhHBTail versusM1-WT) after normalization with the housekeeping gene.

Jin et al.



bio.asm.org/

Dow

nloaded from

mbio.asm.org


absence of a classical protein exporting machinery in GAS andmany Gram-positive pathogens is intriguing, it has also raisedconcerns regarding its physiological relevance. While it is still notclear how this anchorless surface protein is exported to the sur-face, we demonstrate in the present study for the first time thesignificance of the GAPDH export process on the GAS cell surface.This could not have been achieved without employing our previ-ously described novel strategy of creating a GAS mutant express-ing SDH (with a genetically introduced hydrophobic tail at its Cterminus) solely in the cytoplasm or at the cytoplasmic side of themembrane (18). While we clearly demonstrated previously thatthe exportation of SDH to the cell surface is not a mere artifact, inthe present study we were able to unravel the physiological rele-vance of the surface export of SDH by subjecting the M1-SDHHBtail mutant strain to microarray analysis in conjunctionwith various other in vitro and in vivo biochemical/biological as-says.

The major highlights of the present study emerge from twoimportant findings. (i) Simply by retaining SDH in the cytoplasm,carbohydrate metabolism- and many virulence-related genes inGAS were downregulated, and as a result, (ii) the M1-SDHHBtail

mutant was completely attenuated. SDH being a key glycolyticenzyme, its intimate association with GAS carbohydrate metabo-lism is quite obvious. However, its role in the regulation of GAScarbohydrate metabolism could not be established until now ow-ing to the fact that SDH is encoded by a single gene and is indis-pensable for GAS survival. The role of carbohydrate metabolismin GAS pathogenesis was also not evident until carbohydrate uti-lization proteins (CUPs) important in the pathogenesis of GASwere reported (10, 27). It has been demonstrated that many genesencoding these CUPs and especially those that can utilize malto-dextrin and maltose are upregulated when GAS is grown in hu-man saliva (10, 27). Moreover, the downregulation of these genesin the absence of the SptR/S (SPy0874/SPy0875) two-componentsystem indicated that SPy0874 or SptR plays a crucial role in theregulation of complex carbohydrate metabolism in GAS (6). Thepoor ability of the M1-SDHHBtail mutant to grow in the presenceof maltose and maltodextrin concurs with the downregulation of37 out of 38 carbohydrate metabolism genes. However, nochanges in the transcript abundance of the sptR or sptS gene in theM1-SDHHBtail mutant indicate that the cytoplasmically retainedSDH may interfere with SptR/S-mediated functional regulation(e.g., binding to specific promoter) rather than at the sptR/S tran-scriptional level.

Carbohydrate metabolism in bacteria is also regulated by acatabolite control (CcpA) protein (11). Recent reports have eluci-dated the role of CcpA (SPy0514) as a repressor of carbohydratemetabolism and established its association with GAS virulence (5,9, 10, 35, 36). CcpA represses multiple operons involved in utili-zation of nonglucose substrates, including mannose, cellobiose,mannose/fructose, beta-glucosidase PTSs, and also a maltose ABCtransporter (11). Significant downregulation of CcpA-regulatedgenes in the M1-SDHHBtail mutant without altering the expressionlevels of ccpA and hpr indicates that intracellular SDH may poten-tially interact with these proteins to regulate their function. Hpr isthe central regulatory protein of the PTS system that plays animportant role in catabolite control repression (CCR) by its dif-ferential phosphorylation at His (HPr-His~P) and Ser (HPr-SerP) residues. In GAS, the latter binds strongly to catabolite con-trol protein A (CcpA), and the resulting CcpA–HPr-SerP complex

binds strongly (50- to 100-fold higher) to the catabolite responseelement (cre) in the promoter region of CcpA-regulated genes (9,11, 27). It has also been observed that large amounts of doublyphosphorylated HPr [HPr (Ser-P/His~P)] are generated in strep-tococcal and lactococcal species during growth (37). While thesignificance of this dual phosphorylation remains unclear, we ob-served that retention of SDH in the cytoplasm resulted in an in-crease in the doubly phosphorylated Hpr species. This could pos-sibly be attributed to the increased intracellular ATP productionwhich in turn is a consequence of significant downregulation of 14of 25 predicted V-type Na�-ATPase or V1V0ATPase genes(SPy0147-SPy0157 and SPy0754-0760) (38) and arginine deami-nase system (28).

The role of lipid biosynthesis in virulence regulation has beenstudied for only a few Gram-positive pathogens, includingS. pneumoniae and Staphylococcus aureus (39, 40), and there is nosuch information available for GAS. In contrast to the expressionprofile of genes regulating carbohydrate metabolism, upregula-tion of genes responsible for lipid biosynthesis (FAS-II genesSPy1743-SPy1755) prompted us to determine total lipid and fattyacid content. While the saturated fatty acid content increased sig-nificantly in the M1-SDHHBtail mutant, it displayed complete vir-ulence attenuation unlike the phenotype of FAS-II knockout inS. mutans (41). To understand the role of lipid biosynthesis inGAS virulence, the present study invokes future studies on howSDH regulates fatty acid biosynthesis, whether the increased ex-pression of FAS-II genes is the marker of attenuation, and whetherthe biosynthesis of fatty acid plays any role in virulence as in thecase of S. aureus (40).

Although the role of SDH/GAPDH in many nonglycolyticfunctions is directly or indirectly associated with bacterial viru-lence (12, 19, 20, 42), the mechanism of SDH-mediated regulationof GAS/bacterial virulence has so far not been investigated. In thepresent investigation, besides downregulation of carbohydratetransport/metabolism-related genes, we also observed the down-regulation of several virulence-associated genes (e.g., has and sagoperon genes, ska, emm1, collagen-binding protein [Scl], speB, mf,and others). A similar profile for virulence genes was observed forthe M1-GAS mutants (albeit in different type M1 strainMGAS_5005) that lack virulence and carbohydrate gene-regulating two-component systems (TCSs) such as CovR, Mga,and CcpA (5, 27, 43, 44). Interestingly, despite the unaltered ex-pression profiles of the latter genes (covR, mga, or ccpA), thedownregulation of virulence genes in M1-SDHHBtail mutant sug-gests a possible role of SDH in altering the interaction betweenMga and CcpA at the cre site or CovR binding to its promoter siteswhen intracellular equilibrium of SDH is perturbed. The possibil-ity of interaction of SDH with other stand-alone transcriptionfactors regulating GAS pathogenesis and carbohydrate metabo-lism became evident when we observed downregulation of LacD.1(SPy1704 encoding putative tagatose-1,6-bisphosphate aldolase,3.13-fold) and speB gene (7.01-fold) in the SDHHBtail mutant de-spite the fact that SPy1704 is a negative regulator of SpeB. Theexpression of SpeB is also shown to be regulated by ropB/rgg/SPy2042 (45). Significantly reduced SpeB-associated cysteine pro-tease activity with no change in the expression level of SpeB regu-lator ropB/rgg/SPy2042 in the M1-SDHHBtail mutant is additionalevidence for the above-mentioned argument for the ability ofSDH to interact with one or more stand-alone transcription reg-ulator(s).




bio.asm.org/

Dow

nloaded from

mbio.asm.org


To determine whether the export process of SDH per se is re-sponsible for the regulation of virulence, we complementedM1-WT with the gene encoding SDHHBtail and complementedM1-SDHHBtail mutant with the gene encoding SDH. The relativedecrease in the virulence of the M1-WT::sdhHBtail and increase inthe virulence of M1-SDHHBtail::sdh indicated that export of SDH iscertainly essential for GAS to remain virulent, since the retentionof SDH in the cytoplasm results in its attenuation. These resultsalso concur with qRT-PCR-based expression profiles of certainimportant genes studied for these complemented strains (Fig. 6B).In conjunction with these results, the direct binding of SDH toSpeB and the way SpeB (microdomain/ExPortal) and SDH areexported and localized on the cell surface (14, 18, 46, 47) lead us tobelieve that SDH-mediated virulence regulation in GAS couldhappen in three specific ways. (i) SDH by itself can serve as atranscription factor. (ii) SDH can function as an important com-ponent of the transcription complex to regulate the expression ofvirulence-associated genes. (iii) The process of surface export ofSDH by itself regulates the bacterial gene expression profile. SDHmay serve as a chaperone facilitating the export of certain viru-lence factors destined for secretion or exportation.

Together, our findings indicate that the surface exportation ofSDH is an essential phenomenon to maintain GAS virulence. Thisexport is thought to circumvent the imbalance (increased concen-tration) in the intracellular concentration of SDH which other-wise may enhance its multifarious interactions with CovR, SptR,CcpA, Mga, LacD.1, and/or Rgg. While the nature and specificityof these potential interactions are presently unknown, our find-ings open a new avenue of novel regulatory pathways for fine-tuning gene regulation in GAS. Considering the conserved natureof SDH/GAPDH throughout biological evolution and the expor-tation of SDH/GAPDH onto the cell surface in almost all types ofprokaryotes, fungi, and protozoans, we believe that microorgan-isms in general devised this simple export mechanism long beforesophisticated secretion systems came into existence. Thus, expor-tation of SDH/GAPDH on the surface may be an essential func-tion of certain bacteria to remain as successful pathogens. In thiscontext, SDH, being an ancient key glycolytic enzyme, serves as aquintessential regulator fulfilling necessary regulatory functionsof virulence maintenance as and when necessary in a constantlychanging host environment. While the surface exportation ofSDH is indeed an essential process for GAS persistence as a patho-gen, the strategy of preventing this exportation can be exploited asa novel way of attenuation.

MATERIALS AND METHODSBacterial strains and growth conditions. S. pyogenes wild-type strainM1-SF370 (M1-WT; ATCC 700294; American Type Culture Collection,Manassas, VA) and the isogenic mutant strain (M1-SDHHBtail) (18) weregrown in Todd-Hewitt broth (Difco Laboratories) supplemented with0.5% yeast extract (THY) or on proteose peptone no. 3 agar plates (DifcoLaboratories) supplemented with 5% defibrinated sheep blood. Esche-richia coli XL1-Blue was grown in Luria-Bertani (LB) medium (with orwithout agar) at 37°C. Antibiotics were added at the following concentra-tions when needed: ampicillin, 100 �g/ml; and spectinomycin, 500 �g/ml(for S. pyogenes and 50 �g/ml for E. coli). Chemically defined medium(CDM) was prepared as described previously (48), except that 1% glucosewas replaced with other sugars in the same concentrations as carbohy-drate sources, when needed. Unless otherwise mentioned, all the chemi-cals were procured from Sigma-Aldrich. Growth curves of the wild-type

and mutant strains in the presence of various sugar sources were mea-sured by recording the absorbance at 600 nm.

RNA isolation. RNA was extracted from the wild-type (M1-SF370)and mutant (M1-SDHHBtail) strains grown till late exponential phase(OD600 � 0.8), using Qiagen RNeasy kit (Qiagen) as described previously(49). RNase-free DNase I was added to the RNA preparation for removalof residual DNA contamination according to the manufacturer’s proto-col. RNA was quantified, and its integrity was checked by using an Agilent2100 bioanalyzer as described previously (49).

Microarray analysis and real-time PCR (qRT-PCR). Oligonucleotide-based microarray analysis was performed using polylysine-coated slidesspotted with oligonucleotides designed to correspond to 1,769 open readingframes (ORFs) of the S. pyogenes M1-SF370 genome (49). Synthesis and la-beling of the first-strand cDNA, hybridization of probes with microarrays,scanning the signals of bound reagents on the microarray spots, Lowess nor-malization, and R-based microarray analysis and t statistics using Bioconduc-tor multitest (https://carmaweb.genome.tugraz.at/carma/) were performedas described previously (49). A difference of at least 2-fold (log2 ratio of �1 or��1 with a P value of �0.05) in the transcript abundance ratio was consid-ered a significant change in the gene expression. Results were deposited toGEO database and an approved accession number (GSE 15231) was ob-tained. Microarray analysis was validated by real-time quantitative reversetranscription-PCR (qRT-PCR) for randomly selected genes, using brilliantSYBR green QPCR master mix (Roche) employing gene-specific primers (seeTable S1 in the supplemental material) and using a LightCycler480 (Roche)real-time PCR instrument as described previously (49). The copy numbersfor all the genes were normalized with gyrA and proS using two biologicalreplicates each with three technical replicates. The data were analyzed usingthe Exor-4 software (Roche).

Cell fractionation. Cell wall, cell membrane, and cytoplasmic proteinsfrom the wild-type and mutant GAS strains were fractionated by digestingintact GAS cells with recombinant phage lysin (Plys) in phosphate-buffered saline (PBS) buffer containing 30% (wt/vol) raffinose, 5 mMEDTA, and 0.5 mM dithiothreitol (DTT), followed by differential high-speed centrifugation as described previously (12, 50).

Production of recombinant proteins CovR and CcpA and specificpolyclonal antibodies. Recombinant 6�His-tagged CovR (SPy0336) andCcpA (SPy0514) were prepared essentially as described before using pET-14b His-tagged vector and employing nickel-nitrilotriacetic acid (Ni2�-NTA) affinity chromatography (50). The sequences of the specific primersused for cloning the genes in the expression vector are shown in Table S2in the supplemental material. Recombinant proteins were used to raisepolyclonal antibodies in rabbits. Polyclonal antisera were custom-madeby Lampire Biological Laboratories, Inc., using the 50-day express-lineprotocol (Lampire Biological Laboratories, Inc., Pipersville, PA).

Blot overlay and Western blot analysis. Proteins were resolved onSDS-polyacrylamide gels, followed by electroblotting on polyvinylidenedifluoride (PVDF) membranes (Bio-Rad). The membranes were blockedand probed with specific antibody followed by anti-rabbit/mouse/goat(for polyclonal antibody) immunoglobulin G alkaline phosphatase-conjugated secondary antibodies. Immunoreactive bands were visualizedusing either chromogenic (nitroblue tetrazolium [50 �g/ml] and5-bromo-4-chloro-3-indolylphosphate [25 �g/ml]) or chemilumines-cent substrates (Lumi-Phos; Pierce) (19).

Monoclonal antibody 10B6 (51) was used to detect the expression of Mprotein. Sheep antistreptokinase antibody was purchased from AbD SerotecInc. Rabbit anti-HPr was obtained from Joseph Deutscher. Anti-SpeB anti-body was purchased from Santa Cruz Laboratory. Anti-SDH (12), anti-CcpA,and anti-CovR polyclonal antibodies were custom-made by Lampire Biolog-ical Laboratories, Inc., in rabbits injected with purified recombinant proteins.The custom-made antibodies were affinity purified using a proteinG-Sepharose 4B column followed by corresponding purified protein-bound3M phase activated beads as described previously (19).

To determine the binding ability of SDH to SpeB secreted in the cul-ture supernatants of the wild-type, mutant, and complemented strains

Jin et al.



bio.asm.org/

Dow

nloaded from

https://carmaweb.genome.tugraz.at/carma/

mbio.asm.org


grown to late log phase (optical density [OD600] of 0.8), the supernatantswere precipitated with trichloroacetic acid (TCA) (30% final concentra-tion), neutralized, and washed with 100% ethanol containing 1% (vol/vol) M sodium acetate. The proteins present in the precipitates were re-solved on 15% SDS-polyacrylamide gels, Western blotted, and subjectedto the blot overlay method as described previously (19) using a purifiedSDH (12). Briefly, the Western blots were overlaid with purified SDH(10 �g/ml) for 2 h and washed, and the binding of SDH to specific protein(SpeB) was monitored using anti-SDH monoclonal antibody (MAb)13D5 (52) as described above.

Total cell ATP assays. Total ATP content in the wild-type and mutantGAS strains were tested by an ENLITEN ATP assay bioluminescence de-tection kit (Promega) according to the manufacturer’s instructions.Briefly, pellets obtained from 50-ml samples of the GAS cultures grown tolate log phase were suspended in 100 �l of 2 mM EDTA, pH 7.75. The cellswere lysed in the presence of glass beads using FastPrep-120, precipitatedwith 10% trichloroacetic acid in 2 mM EDTA, and quickly centrifuged at16,000 � g for 2 min at 4°C. The supernatant diluted 1:10 (10 �l) was thensubjected to luminescence measurement using FluoStart plate reader. Thetotal ATP was estimated from the standard curve generated using knownconcentrations of ATP.

Total cell fatty acid assays. Total fatty acid content and the composi-tion of the individual fatty acids were custom analyzed by gas chromatog-raphy (Microbial ID Inc., Newark, DE). Briefly, fatty acid methyl esterswere extracted from the GAS cultures grown until late logarithmic phaseand separated by gas chromatography (MIDI). The resulting fatty acidprofiles were analyzed using Sherlock pattern recognition software. Datawere analyzed from three independently grown cultures of the wild-typeand mutant strains. Data were analyzed by one-tailed nonparametric t testwith Welch’s correction using GraphPad Prism 4 software.

TEM. Late-log-phase-grown cultures of the wild-type and M1-SDHHBtail

mutant strains were washed, resuspended in fixative (4% paraformaldehydeand 2.5% gluteraldehyde in 0.1 M cacodylate buffer [pH 7.4]), and subse-quently processed for transmission electron microscopy (TEM) at the OhioState University (OSU) Central Microscopy and Imaging Facility (CMIF)using Technai2 electron microscope as described previously (49, 50).

Measurement of the capsular hyaluronic acid. The capsular hyal-uronic acid from the late-log-phase-grown culture of GAS strains wasextracted with chloroform, and the amount was measured spectrophoto-metrically (640 nm) after the extracts were treated with Sigma Stains-Allsolution as described previously (49, 50).

SLS hemolysis assays. The ability of streptolysin S (SLS) present in theTHY culture supernatants of the late-log-phase-grown (OD at 600 nm[OD600] of 0.8) wild-type and M1-SDHHBtail strains to lyse sheep redblood cells (shRBCs) was measured as described previously (53). Thereleased hemoglobin from the lysed shRBCs was spectrophotometricallymeasured at 570 nm (Bio-Rad). The complete lysis (100% lysis) ofshRBCs was achieved by the addition of 4 M NH4Cl and was used as thepositive control. Samples containing only RBCs and PBS were used asblank or negative control. Percent hemolysis was calculated as follows:[(sample A � blank B)/(100% lysis Control C)] � 100. One hemolyticunit (HU) was defined as the reciprocal of the highest dilution of hemo-lysin that released 50% hemoglobin from 2.5% shRBC solution. To esti-mate the hemolytic activity contributed by streptolysin O, the SLS inhib-itor, trypan blue (50 �g/ml), was added to the samples prior to incubationwith shRBCs.

Measurement of SpeB activity. SpeB-specific cysteine protease activ-ity in the culture supernatant was estimated by visualizing casein hydro-lysis in the form of a clearance zone around the 8-mm 3M paper discsimpregnated with 10 �l of culture supernatants on 5% milk agar plates asdescribed previously (54).

For the quantitative measurement of cysteine protease activity, caseinhydrolysis was analyzed using the fluorescein isothiocyanate (FITC)-labeled casein method (55). For this, filter-sterilized supernatants fromovernight GAS cultures were activated by mixing an equal volume of

activation buffer-SAED (0.1 M sodium acetate buffer [pH 5.0], 1 mMEDTA, 20 mM DTT) and incubating for 30 min at 40°C. The activatedsupernatants were then mixed with an equal volume of FITC-conjugatedcasein (5 �g/ml; AnaSpec Inc.) prepared with 20 �M E-64 or withoutE-64, a cysteine-specific protease inhibitor (obtained from Sigma). Thereaction mixtures were incubated for 1 h at 40°C, and the increase in thefluorescence intensities (excitation [Ex]/emission [Em] � 485/520 nm)was measured by a spectrofluorimeter (Bio-Rad).

In vivo GAS virulence in mice. GAS strains (M1-WT, M1-SDHHBtail,M1-WT::sdhHBtail, and M1-SDHHBtail::sdh) grown until late exponentialphase (OD600 of 0.8) were centrifuged, washed twice with sterile PBS,suspended in the same buffer, and adjusted to an OD600 of 1.0 (represent-ing ~4 � 108 CFU/ml). Five-week-old female CD-1 mice (18 to 20 g, 10 to20 mice/group, Charles River Laboratories) were infected intraperitone-ally with 0.5 ml (~2 � 108 CFU) of the GAS strain and monitored for10 days per the protocol approved by the Ohio State University Institu-tional Animal Care and Use Committee (OSU-IACUC) as described pre-viously (49). Mortality/survival profiles for each infected group were plot-ted and statistically analyzed by the log rank test using GraphPad Prism 4software. After 4 days of infection, the internal organs (spleen, lung, liver,and kidneys) from 4 to 6 sacrificed animals were removed and homoge-nized. GAS strains from the homogenized tissues were recovered on sheepblood agar plates.

Microarray data accession number. Data from eight experimentswere submitted to the Gene Expression Omnibus (GEO) database, as-signed accession number GSE15231, and approved.

ACKNOWLEDGMENTS

We thank Vincent Fischetti (Rockefeller University, New York, NY) for10B6 monoclonal antibody, Joseph Deutscher (INRA, France) for anti-HPr antibody, Daniel Nelson (University of Maryland Biotechnology In-stitute [UMBI], University of Maryland, Bethesda) for recombinant PlyS(lysin), and Haritha Adhikarla for editorial help.

This study was supported in part by the Bill and Melinda GatesFoundation Grand Challenges Explorations Round-3 award(OPP1006841GCE) to V.P. We have no competing financial interests.

SUPPLEMENTAL MATERIALSupplemental material for this article may be found at http://mbio.asm.org/lookup/suppl/doi:10.1128/mBio.00068-11/-/DCSupplemental.

Table S1, PDF file, 0.072 MB.Table S2, PDF file, 0.099 MB.Table S3, PDF file, 0.092 MB.Table S4, PDF file, 0.117 MB.Table S5, PDF file, 0.090 MB.Table S6, PDF file, 0.096 MB.

REFERENCES1. Tart AH, Walker MJ, Musser JM. 2007. New understanding of the group

A Streptococcus pathogenesis cycle. Trends Microbiol. 15:318 –325.2. Carapetis JR, Steer AC, Mulholland EK, Weber M. 2005. The global

burden of group A streptococcal diseases. Lancet Infect. Dis. 5:685– 694.3. Kreikemeyer B, McIver KS, Podbielski A. 2003. Virulence factor regu-

lation and regulatory networks in Streptococcus pyogenes and their impacton pathogen-host interactions. Trends Microbiol. 11:224 –232.

4. Churchward G. 2007. The two faces of Janus: virulence gene regulation byCovR/S in group A streptococci. Mol. Microbiol. 64:34 – 41.

5. McIver KS. 2009. Stand-alone response regulators controlling global vir-ulence networks in Streptococcus pyogenes. Contrib. Microbiol. 16:103–119.

6. Shelburne SA, et al. 2005. Central role of a bacterial two-component generegulatory system of previously unknown function in pathogen persis-tence in human saliva. Proc. Natl. Acad. Sci. U. S. A. 102:16037–16042.

7. Cho KH, Caparon MG. 2005. Patterns of virulence gene expression differbetween biofilm and tissue communities of Streptococcus pyogenes. Mol.Microbiol. 57:1545–1556.

8. Virtaneva K, et al. 2005. Longitudinal analysis of the group A Streptococ-




bio.asm.org/

Dow

nloaded from

http://mbio.asm.org/lookup/suppl/doi:10.1128/mBio.00068-11/-/DCSupplemental

http://mbio.asm.org/lookup/suppl/doi:10.1128/mBio.00068-11/-/DCSupplemental

mbio.asm.org


cus transcriptome in experimental pharyngitis in cynomolgus macaques.Proc. Natl. Acad. Sci. U. S. A. 102:9014 –9019.

9. Almengor AC, Kinkel TL, Day SJ, McIver KS. 2007. The catabolitecontrol protein CcpA binds to Pmga and influences expression of thevirulence regulator Mga in the group A Streptococcus. J. Bacteriol. 189:8405– 8416.

10. Shelburne SA, et al. 2008. A direct link between carbohydrate utilizationand virulence in the major human pathogen group A Streptococcus. Proc.Natl. Acad. Sci. U. S. A. 105:1698 –1703.

11. Deutscher J. 2008. The mechanisms of carbon catabolite repression inbacteria. Curr. Opin. Microbiol. 11:87–93.

12. Pancholi V, Fischetti VA. 1992. A major surface protein on group Astreptococci is a glyceraldehyde-3-phosphate dehydrogenase with multi-ple binding activity. J. Exp. Med. 176:415– 426.

13. Pancholi V. 2001. Multifunctional alpha-enolase: its role in diseases. Cell.Mol. Life Sci. 58:902–920.

14. Pancholi V, Chhatwal GS. 2003. Housekeeping enzymes as virulencefactors for pathogens. Int. J. Med. Microbiol. 293:1–11.

15. Matta SK, Agarwal S, Bhatnagar R. 2010. Surface localized and extracel-lular glyceraldehyde-3-phosphate dehydrogenase of Bacillus anthracis is aplasminogen binding protein. Biochim. Biophys. Acta 1804:2111–2120.

16. Pancholi V, Fischetti VA. 1993. Glyceraldehyde-3-phosphate dehydro-genase on the surface of group A streptococci is also an ADP-ribosylatingenzyme. Proc. Natl. Acad. Sci. U. S. A. 90:8154 – 8158.

17. Lottenberg R, et al. 1992. Cloning, sequence analysis, and expression inEscherichia coli of a streptococcal plasmin receptor. J. Bacteriol. 174:5204 –5210.

18. Boel G, Jin H, Pancholi V. 2005. Inhibition of cell surface export of groupA streptococcal anchorless surface dehydrogenase affects bacterial adher-ence and antiphagocytic properties. Infect. Immun. 73:6237– 6248.

19. Jin H, Song YP, Boel G, Kochar J, Pancholi V. 2005. Group A strepto-coccal surface GAPDH, SDH, recognizes uPAR/CD87 as its receptor onthe human pharyngeal cell and mediates bacterial adherence to host cells.J. Mol. Biol. 350:27– 41.

20. Pancholi V, Fischetti VA. 1997. Regulation of the phosphorylation ofhuman pharyngeal cell proteins by group A streptococcal surface dehy-drogenase (SDH): signal transduction between streptococci and pharyn-geal cells. J. Exp. Med. 186:1633–1643.

21. Alvarez-Dominguez C, et al. 2008. Characterization of a Listeria mono-cytogenes protein interfering with Rab5a. Traffic 9:325–337.

22. Terao Y, Yamaguchi M, Hamada S, Kawabata S. 2006. Multifunctionalglyceraldehyde-3-phosphate dehydrogenase of Streptococcus pyogenesis essential for evasion from neutrophils. J. Biol. Chem. 281:14215–14223.

23. Rosinha GM, Myioshi A, Azevedo V, Splitter GA, Oliveira SC. 2002.Molecular and immunological characterisation of recombinant Brucellaabortus glyceraldehyde-3-phosphate-dehydrogenase, a T- and B-cell reac-tive protein that induces partial protection when co-administered with aninterleukin-12-expressing plasmid in a DNA vaccine formulation. J. Med.Microbiol. 51:661– 671.

24. Liu Y, Oshima S, Kurohara K, Ohnishi K, Kawai K. 2005. Vaccineefficacy of recombinant GAPDH of Edwardsiella tarda against edwardsiel-losis. Microbiol. Immunol. 49:605– 612.

25. Madureira P, et al. 2007. Streptococcus agalactiae GAPDH is a virulence-associated immunomodulatory protein. J. Immunol. 178:1379 –1387.

26. Ferretti JJ, et al. 2001. Complete genome sequence of an M1 strain ofStreptococcus pyogenes. Proc. Natl. Acad. Sci. U. S. A. 98:4658 – 4663.

27. Shelburne SA, Davenport MT, Keith DB, Musser JM. 2008. The role ofcomplex carbohydrate catabolism in the pathogenesis of invasive strepto-cocci. Trends Microbiol. 16:318 –325.

28. Degnan BA, et al. 1998. Inhibition of human peripheral blood mononu-clear cell proliferation by Streptococcus pyogenes cell extract is associatedwith arginine deiminase activity. Infect. Immun. 66:3050 –3058.

29. Marouni MJ, Ziomek E, Sela S. 2003. Influence of group A streptococcalacid glycoprotein on expression of major virulence factors and internal-ization by epithelial cells. Microb. Pathog. 35:63–72.

30. Postma PW, Lengeler JW. 1985. Phosphoenolpyruvate: carbohydratephosphotransferase system of bacteria. Microbiol. Rev. 49:232–269.

31. Thevenot T, Brochu D, Vadeboncoeur C, Hamilton IR. 1995. Regula-tion of ATP-dependent P-(Ser)-HPr formation in Streptococcus mutansand Streptococcus salivarius. J. Bacteriol. 177:2751–2759.

32. Schujman GE, de Mendoza D. 2008. Regulation of type II fatty acidsynthase in gram-positive bacteria. Curr. Opin. Microbiol. 11:148 –152.

33. Datta V, et al. 2005. Mutational analysis of the group A streptococcaloperon encoding streptolysin S and its virulence role in invasive infection.Mol. Microbiol. 56:681– 695.

34. Chiang-Ni C, Wu JJ. 2008. Effects of streptococcal pyrogenic exotoxin Bon pathogenesis of Streptococcus pyogenes. J. Formos. Med. Assoc. 107:677– 685.

35. Shelburne SA, et al. 2010. A combination of independent transcriptionalregulators shapes bacterial virulence gene expression during infection.PLoS Pathog. 6:e1000817.

36 . Shelburne SA, et al. 2006. Maltodextrin utilization plays a key role in theability of group A Streptococcus to colonize the oropharynx. Infect. Immun.74:4605– 4614.

37. Roy DJ, Casabon I, Vaillancourt K, Huot JL, Vadeboncoeur C. 2008.Streptococci and lactococci synthesize large amounts of HPr(Ser-P) (His-P). Can. J. Microbiol. 54:941–949.

38. Takase K, et al. 1994. Sequencing and characterization of the ntp genecluster for vacuolar-type Na(�)-translocation ATPase of Enterococcus hi-rae. J. Biol. Chem. 269:11037–11044.

39. Lu YJ, Rock CO. 2006. Transcriptional regulation of fatty acid biosyn-thesis in Streptococcus pneumoniae. Mol. Microbiol. 59:551–566.

40. Balemans W, et al. 2010. Essentiality of FASII pathway for Staphylococcusaureus. Nature 463:E3–E5.

41. Fozo EM, Scott-Anne K, Koo H, Quivey RG. 2007. Role of unsaturatedfatty acid biosynthesis in virulence of Streptococcus mutans. Infect. Im-mun. 75:1537–1539.

42. Pancholi V, Fontan PA, Jin H. 2003. Plasminogen-mediated group Astreptococcal adherence to and pericellular invasion of human pharyngealcells. Microb. Pathog. 35:293–303.

43. Graham MR, et al. 2002. Virulence control in group A Streptococcusby a two-component gene regulatory system: global expression profil-ing and in vivo infection modeling. Proc. Natl. Acad. Sci. U. S. A.99:13855–13860.

44. Ribardo DA, McIver KS. 2009. Defining the Mga regulon: comparativetranascriptome analysis reveals both direct and indirect regulation by Mgain the group A Streptococcus. Mol. Microbiol. 62:491–508.

45. Dmitriev AV, et al. 2006. The Rgg regulator of Streptococcus pyogenesinfluences utilization of nonglucose carbohydrates, prophage induction,and expression of the NAD-glycohydrolase virulence operon. J. Bacteriol.188:7230 –7241.

46. Rosch J, Caparon M. 2004. A microdomain for protein secretion ingram-positive bacteria. Science 304:1513–1515.

47. Pancholi V, Fischetti VA. 1998. Alpha-enolase, a novel strong plasmin-(ogen) binding protein on the surface of pathogenic streptococci. J. Biol.Chem. 273:14503–14515.

48. van de Rijn I, Kessler RE. 1980. Growth characteristics of group Astreptococci in a new chemically defined medium. Infect. Immun. 27:444 – 448.

49. Pancholi V, Boël G, Jin H. 2010. Streptococcus pyogenes Ser/Thr kinase-regulated cell wall hydrolase is a cell division plane-recognizing and chain-forming virulence factor. J. Biol. Chem. 285:30861–30874.

50. Jin H, Pancholi V. 2006. Identification and biochemical characterizationof a eukaryotic-type serine/threonine kinase and its cognate phosphatasein Streptococcus pyogenes: their biological functions and substrate identi-fication. J. Mol. Biol. 357:1351–1372.

51. Jones KF, Fischetti VA. 1987. Biological and immunochemical identity ofM protein on group G streptococci with M protein on group A strepto-cocci. Infect. Immun. 55:502–506.

52. Fontán PA, Pancholi V, Nociari MM, Fischetti VA. 2000. Antibodies tostreptococcal surface enolase react with human alpha-enolase: implica-tions in poststreptococcal sequelae. J. Infect. Dis. 182:1712–1721.

53. Ravins M, et al. 2000. Characterization of a mouse-passaged, highlyencapsulated variant of group A Streptococcus in in vitro and in vivo stud-ies. J. Infect. Dis. 182:1702–1711.

54. Collin M, Olsén A. 2000. Generation of a mature streptococcal cysteineproteinase is dependent on cell wall-anchored M1 protein. Mol. Micro-biol. 36:1306 –1318.

55. Ma Y, et al. 2006. Identification and characterization of bicistronic speBand prsA gene expression in the group A Streptococcus. J. Bacteriol. 188:7626 –7634.

Jin et al.



bio.asm.org/

Dow

nloaded from

mbio.asm.org


surface export of gapdh/sdh, a glycolytic … glucose molecule. gas metabolizes glucose and derives...

Documents